From 2154c5dcd891cf2a85c0251e07424b5681aa88a2 Mon Sep 17 00:00:00 2001 From: "Haelwenn (lanodan) Monnier" Date: Thu, 18 Oct 2018 07:36:58 +0200 Subject: [PATCH] lib/pleroma/html.ex: Use macros for valid_schemes, change config for schemes --- config/config.exs | 18 +++++++++++++++++- lib/pleroma/formatter.ex | 27 +++------------------------ lib/pleroma/html.ex | 37 ++++++++++++------------------------- 3 files changed, 32 insertions(+), 50 deletions(-) diff --git a/config/config.exs b/config/config.exs index 320296991..e5f0b4f6f 100644 --- a/config/config.exs +++ b/config/config.exs @@ -24,7 +24,23 @@ config :pleroma, Pleroma.Uploaders.S3, config :pleroma, :emoji, shortcode_globs: ["/emoji/custom/**/*.png"] -config :pleroma, :uri_schemes, additionnal_schemes: [] +config :pleroma, :uri_schemes, + valid_schemes: [ + "https", + "http", + "dat", + "dweb", + "gopher", + "ipfs", + "ipns", + "irc", + "ircs", + "magnet", + "mailto", + "mumble", + "ssb", + "xmpp" + ] # Configures the endpoint config :pleroma, Pleroma.Web.Endpoint, diff --git a/lib/pleroma/formatter.ex b/lib/pleroma/formatter.ex index 5b63fb795..d7de5b483 100644 --- a/lib/pleroma/formatter.ex +++ b/lib/pleroma/formatter.ex @@ -171,25 +171,8 @@ defmodule Pleroma.Formatter do @link_regex ~r/[0-9a-z+\-\.]+:[0-9a-z$-_.+!*'(),]+/ui - # IANA got a list https://www.iana.org/assignments/uri-schemes/ but - # Stuff like ipfs isn’t in it - # There is very niche stuff - @uri_schemes [ - "https://", - "http://", - "dat://", - "dweb://", - "gopher://", - "ipfs://", - "ipns://", - "irc:", - "ircs:", - "magnet:", - "mailto:", - "mumble:", - "ssb://", - "xmpp:" - ] + @uri_schemes Application.get_env(:pleroma, :uri_schemes, []) + @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, []) # TODO: make it use something other than @link_regex def html_escape(text, "text/html") do @@ -207,14 +190,10 @@ defmodule Pleroma.Formatter do @doc "changes scheme:... urls to html links" def add_links({subs, text}) do - additionnal_schemes = - Application.get_env(:pleroma, :uri_schemes, []) - |> Keyword.get(:additionnal_schemes, []) - links = text |> String.split([" ", "\t", "
"]) - |> Enum.filter(fn word -> String.starts_with?(word, @uri_schemes ++ additionnal_schemes) end) + |> Enum.filter(fn word -> String.starts_with?(word, @valid_schemes) end) |> Enum.filter(fn word -> Regex.match?(@link_regex, word) end) |> Enum.map(fn url -> {Ecto.UUID.generate(), url} end) |> Enum.sort_by(fn {_, url} -> -String.length(url) end) diff --git a/lib/pleroma/html.ex b/lib/pleroma/html.ex index 8a5ede614..2d2155b82 100644 --- a/lib/pleroma/html.ex +++ b/lib/pleroma/html.ex @@ -3,25 +3,6 @@ defmodule Pleroma.HTML do @markup Application.get_env(:pleroma, :markup) - def valid_schemes() do - [ - "https://", - "http://", - "dat://", - "dweb://", - "gopher://", - "ipfs://", - "ipns://", - "irc:", - "ircs:", - "magnet:", - "mailto:", - "mumble:", - "ssb://", - "xmpp:" - ] - end - defp get_scrubbers(scrubber) when is_atom(scrubber), do: [scrubber] defp get_scrubbers(scrubbers) when is_list(scrubbers), do: scrubbers defp get_scrubbers(_), do: [Pleroma.HTML.Scrubber.Default] @@ -55,6 +36,10 @@ defmodule Pleroma.HTML.Scrubber.TwitterText do paragraphs, breaks and links are allowed through the filter. """ + @markup Application.get_env(:pleroma, :markup) + @uri_schemes Application.get_env(:pleroma, :uri_schemes, []) + @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, []) + require HtmlSanitizeEx.Scrubber.Meta alias HtmlSanitizeEx.Scrubber.Meta @@ -64,7 +49,7 @@ defmodule Pleroma.HTML.Scrubber.TwitterText do Meta.strip_comments() # links - Meta.allow_tag_with_uri_attributes("a", ["href"], HTML.valid_schemes()) + Meta.allow_tag_with_uri_attributes("a", ["href"], @valid_schemes) Meta.allow_tag_with_these_attributes("a", ["name", "title"]) # paragraphs and linebreaks @@ -75,11 +60,10 @@ defmodule Pleroma.HTML.Scrubber.TwitterText do Meta.allow_tag_with_these_attributes("span", []) # allow inline images for custom emoji - @markup Application.get_env(:pleroma, :markup) @allow_inline_images Keyword.get(@markup, :allow_inline_images) if @allow_inline_images do - Meta.allow_tag_with_uri_attributes("img", ["src"], HTML.valid_schemes()) + Meta.allow_tag_with_uri_attributes("img", ["src"], @valid_schemes) Meta.allow_tag_with_these_attributes("img", [ "width", @@ -100,10 +84,14 @@ defmodule Pleroma.HTML.Scrubber.Default do alias Pleroma.HTML + @markup Application.get_env(:pleroma, :markup) + @uri_schemes Application.get_env(:pleroma, :uri_schemes, []) + @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, []) + Meta.remove_cdata_sections_before_scrub() Meta.strip_comments() - Meta.allow_tag_with_uri_attributes("a", ["href"], HTML.valid_schemes()) + Meta.allow_tag_with_uri_attributes("a", ["href"], @valid_schemes) Meta.allow_tag_with_these_attributes("a", ["name", "title"]) Meta.allow_tag_with_these_attributes("b", []) @@ -122,11 +110,10 @@ defmodule Pleroma.HTML.Scrubber.Default do Meta.allow_tag_with_these_attributes("u", []) Meta.allow_tag_with_these_attributes("ul", []) - @markup Application.get_env(:pleroma, :markup) @allow_inline_images Keyword.get(@markup, :allow_inline_images) if @allow_inline_images do - Meta.allow_tag_with_uri_attributes("img", ["src"], HTML.valid_schemes()) + Meta.allow_tag_with_uri_attributes("img", ["src"], @valid_schemes) Meta.allow_tag_with_these_attributes("img", [ "width",